Guidelines for the pharmacological acute treatment of major depression: conflicts with current evidence as demonstrated with the German S3-guidelines

Table 1 Meta-analyses about the efficacy of AD compared to placebo

	Response Rates (at least 50% reduction in depression)
	AD (%)	Placebo (%)	Difference
S3-guidelines summary statement on efficacy these were based on:	50–60	25–35	ca. 25
1. Walsh et al. (2002) [19]	50	30	20
2. Oeljeschläger et al. (2004) [18]^a	67	47	20
Current Meta-Analyses
Cipriani et al. (2018) [20]^b	ca. 50	ca. 40	ca. 10
Jakobsen et al. (2017) [21]^c	49	39	10
Meta-Analyses available before the last update of the S3-guidelines
Furukawa et al. (2016) [23]		35–40
Weitz et al. (2015) [24]	42 (Duloxetine) 45 (SSRIs)	24	18–21
Nelson et al. (2013) [7]	49	40	9
Gibbons et al. (2012) [6]
mild depression	55	37	18
severe depression	58	41	17
Undurraga & Baldessarini (2012) [25]	54	37	17
Melander et al. (2008) (SSRI + SNRI) [26]	48	32	16
Arroll et al. (2005) [27]	SSRI: 56	41	15
Arroll et al. (2005) [27]	TCI: 60	47	13
Storosum et al. (2004) (only TCA) [28]	39	28	11

Notes
^a This review claims a “far-reaching agreement” that two-third respond when treated with AD, whereas there are 20% less responders under placebo, referencing a review of Bauer et al. (2002). The Bauer et al. review, in return, reported a response rate of 50–75% for the old generation AD for medium to severe depression and of 25–33% for placebo (based on a review of the American Psychiatric Association from the year 2000), as well as a response rate of 50% for SSRIs and of 32% for placebo (based on a report from the Agency for Health Care Policy and Research from the year 1999). Thus, the conclusions not only deviate from the cited sources, but these sources are also outdated, since they were published at least 15 years before the publishing of the S3-guidelines
^b Cipriani et al. did not report response rates, but they were estimated elsewhere [29], using an average effect of OR = 1.66 and a response rate of 30–40% for placebo. We also tried to estimate the difference between the AD and placebo response rates, using the results from Jakobsen et al. (2017) [21] who reported 39% responders under placebo. With the average effect of OR = 1.66, we came up with nearly identical results (51% responders under AD and 39% under placebo). Formula: R_AD = OR*R_p/(1-R_p + OR*R_p). R_AD: response rate AD, R_p: response rate placebo
^c based on the results for nonresponse

ISSN: 1471-244X